I’m experiencing exactly the same problem. I have noticed that CPU activity increased considerably after resizing my Droplet (to take advantage of the benefits their new plans).
Recently I opened a ticket to follow up on this situation (so far, they have followed my report.. but there has not been a solution).
In fact.. researching a bit, I found this tweet (so I deduce that I’m not the only one who has this reduction in performance): https://twitter.com/Arajdon/status/955104912907689984
Just before resizing my Droplet, I took a Snapshot from this one.
As a test, I created a new Droplet from that image in same datacenter -SFO2- (I was assigned a new public IP). Once created, I subjected it to an equal load of intensive to the usual and runs perfectly (it was not perceived excessive CPU load).
Once that I proved that my Snapshot worked well and stably, I decided rebuilt my main Droplet (which presents the problem) expecting it to work fine, but I still have the same problem having an excessive CPU load.
As an additional detail, I have noticed that Droplet’s boot takes too long whitout apparent reason (60 secs aproxly, when it regularly was loading in 15 secs at most).
A very practical solution for me would be to create a new Droplet from my Snapshot. The drawback is that I need to keep my public IP assigned initially.